Cuidado Ubíquo de Pacientes com Doenças Crônicas Através de um Modelo de Análise do Comportamento Humano

Abstract

Chronic diseases are among 7 out of the 10 leading causes of death worldwide. The main chronic diseases are heart disease, cancer, chronic respiratory diseases, and diabetes. Heart disease alone causes 9 million deaths a year. Lifestyle changes can prevent many chronic diseases’ deaths and their risk factors. In addition, machine learning and wearable devices have been used for behavior analysis. Therefore, this research proposes B-Track, a computational model for assistance in chronic diseases care through the analysis of behaviors that attenuate or worsen the risk factors associated with chronic diseases, working with user behavior profiles and recommendations for healthier behaviors. The B-Track collects data from different data sources for current and future human behavior analysis through the usage of data fusion and machine learning models. These data comprise the patients’ context histories, which include sensor data and data from self-management surveys. The scientific contribution of B-Track model is the analysis of human behaviors directly associated with risk factors and their susceptibility to the development of NCDs. The model was evaluated through a prototype, which was used within 10 patients during your treatment. Three patients achieved changes in some behaviors over an extended period. Overall, according to the TAM Model evaluation, 83% of users agreed that B-Track was useful, and 80% found it easy to use.